Northumbria Research Linknrl.northumbria.ac.uk/17954/1/does_mitigation_save.pdf · Received 28 April 2014 Received in revised form 12 August 2014 Accepted 13 August 2014 Available

Northumbria Research Link

Citation: Shreve, Cheney and Kelman, Ilan (2014) Does mitigation save? Reviewing cost-benefit analyses of disaster risk reduction. International Journal of Disaster Risk Reduction, 10 (A). pp. 213-235. ISSN 2212-4209

Published by: Elsevier

URL: http://dx.doi.org/10.1016/j.ijdrr.2014.08.004 <http://dx.doi.org/10.1016/j.ijdrr.2014.08.004>

This version was downloaded from Northumbria Research Link: http://nrl.northumbria.ac.uk/17954/

Northumbria University has developed Northumbria Research Link (NRL) to enable users to access the University’s research output. Copyright © and moral rights for items on NRL are retained by the individual author(s) and/or other copyright owners. Single copies of full items can be reproduced, displayed or performed, and given to third parties in any format or medium for personal research or study, educational, or not-for-profit purposes without prior permission or charge, provided the authors, title and full bibliographic details are given, as well as a hyperlink and/or URL to the original metadata page. The content must not be changed in any way. Full items must not be sold commercially in any format or medium without formal permission of the copyright holder. The full policy is available online: http://nrl.northumbria.ac.uk/pol i cies.html

This document may differ from the final, published version of the research and has been made available online in accordance with publisher policies. To read and/or cite from the published version of the research, please visit the publisher’s website (a subscription may be required.)

http://nrl.northumbria.ac.uk/policies.html

Contents lists available at ScienceDirect

International Journal of Disaster Risk Reduction

International Journal of Disaster Risk Reduction 10 (2014) 213–235

http://d2212-42(http://c

n CorrE-m

journal homepage: www.elsevier.com/locate/ijdrr

Does mitigation save? Reviewing cost-benefit analysesof disaster risk reduction

C.M. Shreve a,n, I. Kelman b,c

a University of Northumbria, TACTIC (Tools, Methods and Training for Communities and Society to better prepare for a Crisis) Project,United Kingdomb University College London, Institute for Risk and Disaster Reduction and Institute for Global Health, United Kingdomc Norwegian Institute of International Affairs (NUPI), Norway

a r t i c l e i n f o

Article history:Received 28 April 2014Received in revised form12 August 2014Accepted 13 August 2014Available online 23 August 2014

Keywords:Natural disastersRisk managementEcosystemsCBADRR

x.doi.org/10.1016/j.ijdrr.2014.08.00409/& 2014 The Authors. Published by Elsevireativecommons.org/licenses/by/3.0/).

esponding author.ail address: [email protected] (C.M. Shre

a b s t r a c t

The benefit-cost-ratio (BCR), used in cost-benefit analysis (CBA), is an indicator that attempts tosummarize the overall value for money of a project. Disaster costs continue to rise and thedemand has increased to demonstrate the economic benefit of disaster risk reduction (DRR) topolicy makers. This study compiles and compares original CBA case studies reporting DRR BCRs,without restrictions as to hazard type, location, scale, or other parameters. Many results wereidentified supporting the economic effectiveness of DRR, however, key limitations wereidentified, including a lack of: sensitivity analyses, meta-analyses which critique the literature,consideration of climate change, evaluation of the duration of benefits, broader consideration ofthe process of vulnerability, and potential disbenefits of DRR measures. The studies demon-strate the importance of context for each BCR result. Recommendations are made regardingminimum criteria to consider when conducting DRR CBAs.& 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY

license (http://creativecommons.org/licenses/by/3.0/).

1. Introduction

1.1. Mitigation saves: lives, environment, money

Disaster risk reduction (DRR) has long been recognized inthe literature for its role in mitigating the negative environ-mental, social and economic impacts of natural hazards. Forexample, the US Federal Emergency Management Agency(FEMA), found an average benefit-cost ratio (BCR) of 4 in areview of investments in 4000 mitigation programs in the US[63,54]. Still, DRR benefits are largely under-quantified incomparison to the frequency of disasters and the resultingimpacts, especially in developing nations [54]. For example,for flood mitigation in Mozambique, the post-disaster aidrequest was 203 times the unfulfilled pre-disaster request[55].

er Ltd. This is an open acces

ve).

Additionally, myths have arisen surrounding BCRs for DRR.The most infamous is the often-quoted ratio that the WorldBank is purported to have calculated that DRR saves $7(sometimes $4–7) for every $1 invested. The 7:1 ratio con-tinues to be used today, often without citing a reference, forexample, by top UN officials [80], government organizations(USAID, e.g. [3]), and NGOs (Center for American Progress, e.g.[57]; Oxfam, e.g. [68]). The World Bank no longer promotesthat specific statement and recommends that the ratio not beused (Kull, personal communication). The origins of this ratiocould not be tracked down, with the earliest citation found sofar being [13] stating, without a source, that ‘The World Bankand U.S. Geological Survey calculate that a predicted $400billion in economic losses from natural disasters over the1990s could be reduced by $280 billion with a $40 billioninvestment in prevention, mitigation and preparedness stra-tegies’. When each author was contacted, given the length oftime that had elapsed since Dilley and Heyman [13] waspublished, it was difficult for either to provide moreinformation.

s article under the CC BY license

www.sciencedirect.com/science/journal/22124209

www.elsevier.com/locate/ijdrr

http://dx.doi.org/10.1016/j.ijdrr.2014.08.004



http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijdrr.2014.08.004&domain=pdf



mailto:[email protected]


C.M. Shreve, I. Kelman / International Journal of Disaster Risk Reduction 10 (2014) 213–235214

It is also important to note that DRR does not inevitablyor necessarily have a favorable BCR, as noted in somestudies analyzed throughout this paper. There is also thequestion about whether or not a hazard must manifest forthe BCR to be appreciated. For instance, if flood riskreduction measures are taken inside a property but noflood manifests over the lifetime of that building, are thebenefits of the measures accrued and was it worthwhile totake the measures? These risk management discussionsare limited in the studies. More could also be discussedregarding co-benefits of DRR measures, so that meas-ures undertaken yield gains irrespective of a hazardmanifesting.

Nevertheless, as disaster costs continue to rise and aspolitics continues to shift towards justifying actions infinancial terms, the demand has increased to demonstratethe economic benefit of DRR to policy makers and decisionmakers [17,2,40,27,53]. If the financial benefits can beshown, a stronger possibility exists for investment indisaster mitigation actions, although that is by no meanscertain.

Yet, for example, despite FEMA's work [63,54], in theU.S., only 10% of earthquake- and flood-prone householdshave adopted mitigation strategies [54]. That despitefloods from Hurricane Katrina (2005) and Hurricane Sandy(2012) each costing more than $100 billion—with a similarfigure expected as the cost of the next major U.S. earth-quake whether that strikes Los Angeles, St. Louis, orBoston. Meanwhile, studies cover a wide range of para-meters in terms of locations, DRR measures, hazards, andtemporal scales, including approaches which might notalways be considered as core DRR activities even thoughthey are and should be central to DRR.

For example, Kull [52] utilize a ‘people-centered’resilience-driven flood risk reduction approach in Indiafinding greater economic efficiency, lower initial invest-ment costs, and returns that are not sensitive to assump-tions traditionally made during CBA (e.g. discount rates,future climate conditions) when compared to structuralflood mitigation measures in the region. Khan [47] demon-strates technology interventions, such as a new boat winchsystem in Vietnam. The Red Cross (2008) presents one of afew examples of evaluating the benefits of training withthe inclusion of First Aid training in its CBA for its work inNepal. Mechler [62] and Kull [52,53] include climatechange scenarios in their CBAs, perhaps providing a morecomprehensive projection of potential costs. Dedeurwaer-dere [12], UNIDSR (2002), and Nepal Red Cross [64]evaluate ecosystem restoration approaches such as refor-estation of mangroves and rain forests, which contri-bute to sustainable livelihoods, ecosystem stability, andreduce risk.

The plethora of studies on, and the concern about,disaster costs has led to studies compiling this informa-tion. For example the global and multi-peril databasesgenerated by Munich RE and CRED (the EM-DAT database)span space, time, and hazard types. The equivalentapproach for DRR benefits does not exist. This paper is astart towards setting up a framework for comparing DRRBCRs across multiple case studies in space, in time, and fordifferent hazards and vulnerability characteristics.

2. Methods and questions

Cost benefit analysis (CBA) is an established economictool for comparing the benefits and costs of a given projector activity [50,2,18,82,53]. CBA consist of four primarystages: (i) project definition, in which the reallocation ofresources being proposed are identified (ii) identificationof project impacts, including assessment of additionality(net project benefits) and displacement (‘crowding-out’),(iii) evaluating which impacts are economically relevant,that is, quantifying the physical impacts of the project and(iv) calculating a monetary valuation, discounting, weight-ing and sensitivity analysis [26]).

As Venton [82] and many other studies demonstrate, theutility of CBA extends beyond a tool for cost comparison todecision support. Referring to an Oxfam study undertaken inEl Salvador in 2010, Venton [82] reflects on the finding thatthe use of community-based silos and storage practices toprotect crops were not actually cost-effective, in large part dueto cultural barriers to collective storage that dictated the need(and expense) of individual household silos. CBA was instru-mental in this case in evaluating alternative measures, betterenabling a discussion between community based organiza-tions (CBOs) and the government to find a culturally accep-table and cost-efficient solution.

CBA has limitations that are recognized, some of whichare inherent to every analysis. For example, for environ-mental issues, (i) technical limitations for the valuationof non-market goods, such as wildlife or landscapes,(ii) inability to predict what project impacts will be onecosystems, (iii) lack of methods for incorporating uncer-tainty and irreversibility [26]). Other frequent criticisms ofCBA for DRR and other purposes are a lack of quantificationof the distributional impacts (e.g. who benefits and whopays?) [52], ethical concerns over associating a monetaryvalue to life [60], and quantifying other intangibles [54].More contextually, CBAs for DRR tend not to quantify socialand environmental impacts, while some of these benefitsare qualitative and therefore are not quantifiable with CBA—or even comparable in terms of costs and benefits.

Despite these limitations, CBA is still a commonly reliedupon metric for communicating benefits to decisionmakers. CBA can be used to formulate economic argu-ments for investment in risk reduction, rather thanresponding to the impacts of a future disaster event [82].In terms of specific components of the CBA, the benefit-cost-ratio (BCR) is an indicator used to summarize theoverall value for money of a specific project.

The examples of CBA for DRR cited above range acrosshazard types, geographies, scales, and vulnerabilities. Thesestudies rarely report the costs and benefits of these DRRstrategies in a systematic manner to facilitate an understand-ing of which technique might be best in which circumstance.

This study compiles and compares original CBA casestudies reporting DRR BCRs, without restrictions as to hazardtype, location, scale, or other parameters. To be included here,a study must provide a new, quantitative BCR for a DRRinitiative, indicating the savings obtained for the investment.Only studies reporting such numbers, and the methodologiesand data used to obtain the ratio, are included. For instance,studies only describing methods or without full data analysis

Table 1Descriptions of DRR activities, benefits, costs and main study parameters.

Authors Targetbenefactors

Level Hazard(s) DRR activitiesevaluated

DRR activitybenefits

Vulnerability:valued items(description)

Vulnerability:items not valued,rationale (whereprovided)

Time frame Discountrate

Cost-benefit(C/B) orbenefit-cost (B/C)

Structural,non-structural

Framing

Venton[82]a

Agricultural–pastoralists inMzimbaDistrict,Malawi

Communitybased

Drought Provision ofalternative croptypes and early-maturing seedvarieties;donation of 2breeding goats toeach household;training in soilwaterconservation(swc);contingencyplanning forfuture shocks

Improved cropyields;increasedlivestocknumbers;increased use ofswc techniques(e.g. Water-harvesting andmicro-irrigation)

Maize yield andnumber of goats perhousehold; loss ofeducation and laboravoided ; proxy forloss of life avoided, e.g. Earnings thatadults would havemade if alive wereestimated)

Any indirect impacts 10-yrs 0, 10% (B/C) 24 to 35(for 10%, 0%discount rates,respectively)

Non-structural

Backward-looking

KhogaliandZewdu[48]

(1) Pastoralistsforced intosemi-permanentresettlementsin Al Manaar,Derudeib; (2)agricultural-pastoralists inLashob; (3)households inthe Hamisietregion; (4)water fornomadicpastoralistsand theirlivestock whomigrateannually

Communitybased

Drought (1) Constructionof terraces; (2)construction ofearthembankments;(3)CommunalVegetable Garden(irrigated); (4)hafir construction(large hole dug inthe ground thatstores runoffwater)

(1) Householdsable to producesorghum thatwere previouslynot able to; (2)sorghumproductionduring drought;(3) sorghumand vegetableproduction; (4)reduced deathand improvedhealth oflivestock;reduced conflict

(1) Increasedproduction capacity:sorghum,vegetables,livestock; (2)construction,materials, training,labor, seeds,maintenance; (3)number ofhouseholds in thearea benefitting fromproject; value ofsorghum; (4)livestock, wages lostfrom inability towork

(1–3) Land is not soldin region and has nomarket value; (4)cost of maintenancefor embankments(made of soil); water;market prices

10-yrs (1–3);15-yrs (4)

10% (C/B) (1) 1: 61;(2) 1: 2.4; (3)1:1800; (4)1:2.7

Structuraland non-structural

Assessesbenefits fromdifferent DRRprogramactivities

Mechler[62]

Residents indrought proneUttar Pradesh,India

Communitybased

Drought (i) Subsidizedmicro-cropinsurance forspreadingdrought risk,development of(ii) groundwaterirrigation and (iii)a combinationof i, ii

Reducedincome by thefarmer fromdiversionactivities,reduced reliefexpenditure

Groundwaterirrigation, boreholeconstruction,pumping water,insurance premiumsand technicalassistance

Other social benefitsand benefits tobroader societalgroups; out of scopeof project, as itconsidered a certaindemographic(vulnerable, poorfarmers)

43 yrs(2007–2050)

0–20% (B/C) 1–3.5 Non-structural

Forward-looking

Khan [47] Residentsvulnerable toearthquakes in

Communitybased

Earthquake Utilizing straw-bale in buildingconstruction

Reduced priceof buildingmaterials,

Building materials,maintenance andcost of reduction in

Human life (ethicalimplications)

30 years 12% (C/B) 2.0 Structural Ex-ante(forecastbased)

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

215

Table 1 (continued )









Framing

Kathmandu,Nepal

instead of brickconstruction

reducedheating/coolingcosts, straw-bale structuresare resistant toearthquakes(reduced liveslost), decreasein child labor(common forbrickconstruction),improved airquality

floor areanecessitated by thewider straw-baleconstruction for atypical 2 story houseKathmandu,decreased healthcosts

Kunreu-therandMichel-Kerjan[54]a

Students andschool staff in35 of the mostseismicallyactivedevelopingcountries

Nationalstudy

Earthquake Retrofittingschools in 35seismically activecountries in thedeveloping worldso they areearthquakeresistant

Over the next50 years anestimated250,000 livescould be savedin 35 countrieswith aninvestment of$300 billion toretrofit schools.As the value oflife (human life) component isincreased in theanalysis theBCRs increase,e.g. for a humanlife of $1.5 M, 13countries havea BCR41, $75billion could bespent onretrofittingschools andmore than135,000 livescould be saved

Human life, cost ofretrofitting schools(construction)

Social,environmentalbenefits (beyondscope of study)

10-, 25-, 50-yrs 5, 12% (B/C) as value oflife increases,BC exceeds onefor manycountries forretrofittingschools (e.g. at3% discountrate, BC exceeds1 for 13 of the35 countriesand 135,000lives could besaved over thenext 50 yrs)

Structural Evaluates thecosts andbenefits ofalternativeprograms andpolicies forreeducatingfuture damagesand fatalitiesfrom naturalhazards andfacilitatingrecovery

Holland[33]

Residents inNavua, Fiji

Communitybased

Flood Early warningsystem

Decreasedeconomic lossfrom: reducedinjury (peoplehave warning/more time toevacuate),personal and

Economic lossestotaled fromhousehold losses(homes, premises,possessions),business losses,payment fromgovernment, NGOs,

Certain humanitarianaid items, traumaand irreplaceableitems, days lost forschool children dueto water shortages

20-yrs 3, 7, 10% (B/C) 1,7 (forgovernment,internationalstakeholders,respectively)

Non-structural

Assessesimpacts acrosssectors anddistributionalissues (as citedin [82])

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

216

commerciallosses (peoplehave more timeto movevaluables),reduced aidfromgovernmentand othersources (peoplecan betterprotectpossessions)

charityorganizations, otherlosses (trauma/medical).

EWASE[16]

Communitiesin flood proneregions ofAustria

Communitybased

Flood Effectiveness ofearly warningsystems in smallriver basins thathave shorthydrologicalresponse timescompared to thecost of structuralflood measures

Increase in leadtime mayprovidevaluable timefor completionof preventativemeasures;however, a falsealarm will haveeconomic costs

Early warning system(investment costs,maintenance andphysical assets andmaintenance, andoperating costs)

Not included in CBAwere intangibledamages, but theseare addressedseparately in a multi-criteria assessment

20-yrs 3% (B/C) (earlywarningsystem) 2.6–9.0

Non-structural

Assessespotentialeconomicbenefits ofearly warningsystem/meteorologicalservices versuscosts of earlywarningsystem/meteorologicalservices

HolubandFuc-hs [34]

Local buildings/infrastructurein AustrianAlps

Communitybased

Flood Local structuralmeasures

Preventeddamage tobuildings/infrastructurein study site

Potential damage tobuildings from flashfloods; cost of localstructural measures

Downstreambenefits; value ofitems withinbuildings

80-yrs 3.5%(interestrate)

2.1–6.7 Structural Comparativeanalysis ofmitigationstudies

Mechler[61]b

Piura, Peruresidents inflood pronearea

Communitybased

Flood Polderconstruction

Elevatingexisting dykesand addingpoldersdecreasesflooding risk

Private sector:housing damaged ordestroyed; educationand health, waterand sewage,agricultural, industry,commerce andservice sectors:assets destroyed ordamaged (buildings,machinery, roads,etc.)

Environmentaldamage (no data)and environmentalbenefits (e.g.increasedreforestation due toincreased rainfall)

30-yrs 12% (B/C) 2.2–3.8 Structural Backward-looking

Alsopub-lishedinMechler[61]c

Semerang,Indonesiaresidents inflood pronearea

Communitybased

Flood Return on anintegrated watermanagement andflood protectionscheme (e.g.reducing groundsubsidence bydecreasinggroundwaterwithdrawal),improveddrainage tomitigate tidalinundation)

Reducedflooding andinundation

Construction andoperation costs forstructural mitigationmeasures

Broader socialbenefits not included

54-yrs (2005–2059)

12% (B/C) 1.9–2.5 Structural Forward-looking

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

217










Framing

BurtonandVenton[5]

Residents inPhilippinesunder naturalhazard threatwhere DRRprograms areimplemented

Communitybased

Flood Cost-benefitanalysis of theIntegrated BasedDisasterPreparednessprogram (ICBDP)versus disasterresponseoperationsundertaken bythe PhilippinesNational RedCross

The protectionof assets suchas housing,crops andlivestock;health benefitssuch as accessto safe waterand socialbenefits such asthe safe accessof children totheir schools

Construction cost ofstructural measures(hanging footbridge,sea wall, dyke)

Authors note thatconsiderable datalimitations limit theCBA to only lookingat some of the small-scale physicalmitigation projectsundertaken throughthe CBDRM program

15-yrs Notspecified

(B/C) 24(footbridge);4.9 (sea wall);0.7 (dyke)

Structural Backward-looking

White andRorick[84]

Residents inflood proneKailali, Nepalparticipating inDRR program

Communitybased

Flood Multi-sectoredand relies on amix of capacitybuilding, physicaland early warninginterventions (e.g.bio-engineeringfor riverbankprotection such asbamboo cribwalls, plantationson the river bank,evacuation routes,boats, raisedwater points,embankmentwork and spurs,early warningsystems,communityplanning andcapacity building)

Reducednumber ofhouses flooded,reduction ingrain storagelost, asset lossin floodedhomes avoided,crops were stilllost, percentageof land lost toerosiondecreased,infrastructureloss remainedsame, numberof individualsexposed tocontaminateddrinking wateravoided

Damage to housesflooded/assets inhouses; grain-storage and annualcrops lost, landpermanently lost dueto erosion,infrastructure lost,number ofindividuals exposedto contaminateddrinking water;household sizes/value of land owned

Qualitative social andenvironmentalbenefits were notmonetized

10-yrs 10% (B/C) 3.49 Structuraland non-structural

Backward-looking

[64] Residents inIlam District,Nepalexperiencingflood hazards

Communitybased

flood Mitigation works(construction offlood containingwalls, gabionboxes built in theriver bed, treeplanting onriverbanks),maintenance oftube wells,construction ofevacuation

Householdsborrow moneyat 2% rate; land/crops protectedby mitigationworks;livestockbrought to safeareas duringhazards due topreparednessplans; wells

Land, crops, housesprotected bymitigation works;income generationloans; protection ofwater sources; firstaid training

Livestock protected(minimal impact),greater protection offorest resources(outside of studyscope), provision ofshelter/relief items(outside of scope ofstudy), social impactssuch as improvedcoordination,empowerment of

15-yrs 10% (B/C) 18.6(sensitivityanalysis 14.8)


Backward-looking

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

218

shelters,formation ofcommunity DRRunits, emergencyfund, first aidtraining, supply ofa rickshawambulance

protected fromcontamination;houses stilldestroyed byfires/elephantattacks, butemergencyfund (cash andgrain) providessecurity forthose affected;reduced casesof diarrhea/illness,rickshawprovides fastervisit to doctor

women, greatersense of security(cannot assignquantitative value)

[28] Residents inDez and KarunRiverfloodplains,Iran

Communitybased

Flood Structuralmitigationmeasuresincluding dykes,levees, floodretention damsand flooddiversion

Avoided orreduced flooddamages

Construction costs Social,environmental costs(outside of projectscope)

25-yrs 10% (B/C) 0.29–1.03levees, 0.7–1.34dams, 1.1 flooddiversion

Structural Backward-looking andforward-looking

Khan [46] Residents inflood pronearea of LaiBasin, Pakistan

Communitybased

Flood (1) Expressway/channel; riverimprovements;(2) early warningsystem; (3)relocation ofhouses alongflood plain andrestoration of areawith wetland

(1) Highwaysmore floodresistant;reduced peakriver flow andincreased flowcapacity due toriverimprovements;(2) decreaserisk of injuryand loss of lifefrom flooding,reduceddamage toproperty ifresidents havesufficient timeto takeprecautionarymeasures; (3)reduce oreliminate riskof householdspreviously inthe floodplain,ecologicalimprovementsthroughrestoration

(Vulnerability) usingrisk and damage datafrom 2001 flood andtriangulation ofproperty valuesconducted with realestate agents in thefloodplain; (depthdamage) data fromvarious regional andglobal studies of theregion, corroboratedwith anecdotalevidence andqualitative surveys ofthe area; (economiceffects) reporteddamage from 2001flood; reportedillness from malaria

Social benefits offlood prevention (e.g.reduced diseaseburden, trauma,disruption oflivelihoods) notincluded becausethere was no reliabledata

30-yrs 12% (B/C) (1) 8.55–9.25; (2) 0.96;(3) 1.34


Backward-looking

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

219










Framing

Kull[52,53]c

Residents inflood proneGangetic Basin,Nepal and India

Communitybased

Flood Individual level(raising houseplinths andfodder storageunits, rainwaterharvesting,raising existingprivate handpumps andtoilets);community level(early warningsystem, raisingcommunity handpumps andtoilets,constructing floodshelters,establishing grainand seed banks,maintenance ofkey drainagebottlenecks,development ofself help groupsand purchasingcommunityboats); societylevel (promotionof flood adaptedagriculture andstrengthening ofhealthcaresystem).

Reduce risk ofdeath, injury orillness relatedto flooding;improveagriculturalpractices andproductivity

Survey questionnairecollected informationon specific disaster-related loss, coping,exposure,vulnerability,preference and cost/benefit data; cost of2003 embankmentproject

Authors note that,while conclusionsappear robust, dataavailability andquality stillconstrained theanalysis

43 yrs (2007–2050)

0–20% (B/C) 2–2.5 Structuraland non-structural

Backward-looking(Nepal),backward-looking andforward-looking (In)

IFRC [39] Residents inflood pronecommunities ofBangladesh

Communitybased

Flood Creation ofcommunitygroups to raiserisk awarenessand increasepreparedness,construction ofescape routes,set-up ofcommunitydisasteremergency funds,construction of

Communitygroups raisehazardawareness, aswell as healthand sanitationknowledge;evacuationroutes decreaseloss of life andinjury;emergencyfund allows for

Household surveysand reports wereutilized to estimateDRR program costsand benefits

Improvedcommunitycohesion/greatersense of security;lives saved, injuriesavoided; hybridvegetable seeds(future benefits), etc.

15-yrs 7.74% (B/C) 1.18–3.04;futureprotectivebenefits (3.05–4.90)


Cba conductedto assesseconomicefficiency of drrprograms

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

220

tube-wells toincrease access todrinking water;training andawareness raisingin health andsanitation

rebuilding afterhazard events

Kunreu-therandMichel-Kerjan[54]

Residents in 34countries mostprone to flooddamage

Nationalstudy

Flood Constructing aone-meter highwall to protecthomes in 34 ofthe most floodpronecommunitiesglobal

For aninvestment of$904 billion inconstructingwalls aroundhouses or $.5.2trillion toelevate housesin 34 of theworst floodimpactedcountries61,000 livesover the next50 years couldbe saved

Expected reductionsin damage toinfrastructure,property and avoidedfatalities

Other social,environmentalbenefits (beyondscope of study)

10-, 25-, 50-yrs 5, 12% (B/C) 60building one-meter wall;14.5 forelevatinghomes

Structural Forward-looking

VentonandVenton[81]

Residents intwo drought/flood pronecommunities inIn where DRRactivities havebeenimplemented

Communitybased

Flood, drought Two communitieswith existing DRRprograms wereselected toevaluate thebenefit of the DRRactivities; impactswere analyzed infive categories(natural, physical,human, social andeconomic)

(Bihar) plantingof trees toincrease soilstability; villagedevelopmentfund able toprovide loansfor rebuilding;raised handpumps ensureclean watersupply; reducedlosses andinjury to peopledue to effectiveevacuation;provision ofboats meanscommunitydoes not haveto rent;(Khammam)raised handpumps ensurewater supply,reduce healthproblems, andensure noblockage oncefloodwatersrecede;provision ofhand pumpsallows for

(Bihar) installation ofhand pumps, boats,motorbike for stafftransport,construction ofevacuation road,community training,personnel supportcosts (office rental,travel/lodging,stationary/printing,communication),personnel costs(project staff andconsultancy)

Not valued becauseof lack of data:destruction of crops/soil from severefloods/drought;houses destroyed infloods; health costsof flood/drought;social relationshipcosts; health ofsurvivors

20-yrs (Bihar);15-yrs(Khammam)

10% (B/C) Bihar(baselinescenario 3.17–4.58), raisedhand pump:3.2, potentialfutureinitiatives: 0.62,low interestloans: 57.8);Khammam (B/C) (baselinescenario 3.7–20.05)

Non-structural

Backward-looking

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

221










Framing

easier access towater, enablinglivelihoodactivities andeasier irrigationof fields

Dedeur-waer-dere[12]

Residents inPhilippinesimpacted byfloods andlahars

Communitybased

Flood, lahar Rainforestationfarming (15 yrs);bambooplantation (10 sqkm, 4 yrs); riverchannelimprovements (3yrs);

Reduction ineconomic lossesfrom flooding/lahars

Potential economiclosses (PELs) werebased on theeconomic values ofthe existinginvestmentper sector, e.g.agriculture (mainlycrops); Properties(industry and privateinvestments);Infrastructure (roads,bridges and the like).Benefits of thenatural disastermanagement arethen measured as thedifference betweenPEL without and withthe project.

Environmental orsocial benefits, nodata

3-yrs (Riverchannelimprovement);4-yrs (bambooplantation); 30-yrs(rainforestationfarming)

12% (B/C)Rainforestation30; Bambooplantation14.74; riverchannelimprovements3.5 (dredging,dikeconstruction,widening andchannelexcavation/dredging,construction offloodways)


Forward-looking

MMC[63]d,[20,72,-22,85]

Selectedcommunitiesandrepresentativenationalsample, (USA)

Nationalstudy

Hydro-meteorological(general)

Varied bycommunity

Loss avoided:propertydamage (e.g.buildingscontents,bridges,pipelines),direct businessinterruptionloss (e.g.damagedindustrial,commercial orretail facilities),indirectbusinessinterruptionloss (e.g.ordinarymultiplier or‘ripple’ effect),

Supplementalmethods were usedto assess directproperty losses fromfloods andtornadoes; casualtylosses fromhurricanes,tornadoes andfloods; businessinterruption lossesfor utilities;environmental andhistoric preservationbenefits; and processmitigation activities;project cost data

Uncertainty inmodels/database andheterogeneity ofcommunities; ‘rippleeffects’

1998–2005(Communities);1993–2003 fornational study

2% (B/C) 3.5–5.1;average of 4acrossprograms

Structuraland non-structural(dependingon type ofgrant)

([63] andSupportingstudies)assessing thefuture savingsfrommitigationactivities

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

222

nonmarketdamage (e.g.environmentaldamage towetlands,parks, andwildlife anddamage tohistoricstructures),human losses(e.g. deaths,injuries,homelessness),cost ofemergencyresponse (e.g.ambulanceservice, fireprotection)

GuocaiandWang[27]

End-users ofmeteorologicalservices inChina

Nationalstudy


Meteorologicalservices, dividedinto public andfor variouseconomic sectors

Economicbenefits gainedthrough(public)utilizingweather servicefor planningand avoidinglosses;(government,business)disasterplanning

Survey methodevaluatesparticipantseconomic benefit viaI) willing-to-pay, II)cost-savings, and III)shadow-pricemethods

Accuracy of themeteorologicalservices; additionalcosts of the service(TV, radio, internet);uncertainty

Not specified notspecified

(C/B) 1:40 Non-structural

Backward-looking

NOAA[66]

End-users ofGOES-Rmeteorologicalproducts inaviation,energy(electricity andnatural gas),irrigatedagriculture,andrecreationalboating

Nationalstudy


Improvedmeteorologicalforecasts utilizingGOES satellites

Improvedtropical cycloneforecasting(more effectiveaction toprotectproperty andenableevacuation);enhancedaviationforecasting(improvementsin avoidabledelays, value ofpassenger timeavoided,avoidablephysical assetsandmaintenancecosts, andavoidable riskof aircraft/life

Costs associated withimprovingmeteorologicalprogram (e.g.technology,infrastructure,program costs)

Other potentialbenefits of the GOES-R satellite were notvalued, e.g. benefitsto human health(monitoring ofharmful events suchas algae blooms andforest fires);monitoring of waterquality, river flowsand reservoirmanagement,monitoring of oceanresources (seasurface temperaturenear corals, oceancurrent monitoring)were not valued

12-yrs (2015–2027)

7% note: CBAresults notpresented inratios; dollaramounts ofestimatedsavings given.

Non-structural

Forward-looking

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

223










Framing

lost); moreaccuratetemperatureforecasts(improvedenergy demandexpectationsand savings inelectricity/natural gassectors);enhancedforecasts lead(more efficientirrigation ofcrops)

Khan [47] Fishermenimpacted bysevere weatherin Vietnam

Communitybased


Installation of aboat-winchsystem

Money savedon fuel cost,safety (no oneis required tostay on ship),time (boats arepulled intoshore faster);less disruptionto livelihoodsbecause waittimes arereduced

Sunken boats andships, damaged boatsand ships, cost oflivelihood disruptionfrom false alarms,damaged houses

‘Peace of mind’knowing that theywould not have towait long hours fortheir boats to behauled ashore; thiscannot be valued

30-yrs 12% (B/C) 3.5 Non-structural

Backward-looking

IFRC [38] Vietnamresidents in ornear coastalafforestationprograms

Communitybased


Mangroveafforestationalong coastline(for wave-damping actionplus habitatbenefits forfishes/fisheries);bamboo plantingbetween riverbanks and dykes;tree plantingalong coastline(for wind-breakingcapabilities)

Protectivebenefits ofmangroves(reduced costsin: sea-dykemaintenance,disaster-inducedmaterial losses(publicinfrastructure,buildings,crops, livestock,aquaculture)and non-material losses(injuries,death), indirect(long-term)

Protection fees,planting costs(community wagefees)

Wider ecologicalbenefits; dataavailability

31-yrs (1994–2025)

7.23% (B/C) 18.64–68.92(depending onafforestationactivities indifferentcommunities)

Non-structural

Forward-looking

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

224

losses (e.g.reducedproductivitydue to saltwaterintrusion orinjuries),shorelinestabilization);economicbenefits(planters’income,increased yieldfrom collectionof animals oranimalproducts orwoodcollection),ecologicalbenefits(carbon value,nutrientretention,sedimentretention,biodiversityhabitat)

WorldBank[90]

End-users ofnationalmeteorologicalservices inBelarus,Georgia andKazakhstan

Nationalstudy


Proposedmodernization ofthe nationalmeteorologicalservices (e.g.improving status/capabilities anddelivery)

Avoidedeconomic lossfrom naturalhazards

Damages incurredfromhydrometeorologicalhazards, e.g. toagriculture,communal services,transport andcommunication,additional costs ofirrigation, energy

Does not considerlosses resulting fromless-than-criticalhydrometeorologicalphenomena, i.e.,those that are notclassified asemergencies; asidefrom this, thestatistics do notcover all aspects ofweather impact onthe economy

3–5 yrs 10% (B/C) 3.3(Belarus); 5.7(Georgia); 3.1(Kazakhstan)

Non-structural

Forward-looking

VentonandVenton[81]e

Islanders in 3Maldivianislands wherecyclones andsevere weatherare a concern

Communitybased


Based on SafeIsland Protection(SIP) plan, whichis not explicitlydetailed in thetext, but includesboth structuraland non-structuralmitigationmeasures

Vilufushirehabilitatedafter tsunami toa ‘safer island’;decreaseddamage fromsevere weather

Hazard assessmentand impact fromDIRAM1/2 databases;climate changeassessment primarilyliterature based;probability ofhazards waspresented as a range(lack of data);‘income over life’method (for liveslost); proxy valuesfor economic value of

Does not include thewider impacts of asafer island program,such as costs ofrelocation, decreasedinfrastructure costson ‘abandonedislands’, or macrolevel impacts to GDPbecause the studyfocuses on theislands themselves

50-yrs 0–15% Thinadhoo(SIP): 0.39–1.40, selectedSIP: 0.52–1.85,limitedprotection:1.13–3.65;Viligili (SIP):0.28–1, selectedSIP: 0.29–0.96,limitedprotection:0.42–1.33);


Forward-looking(Thindahoo,Vigili),backward-looking(Vilufushi)

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

225










Framing

Vilufushi, as theisland was destroyedduring a tsunami

Vilufushi (SIP:0.50–1.95)

[54] Residents in 34countries mostprone to winddamage

Nationalstudy


DRR measures toimprove roofprotection inhurricane andcyclone proneregions

$951 Billion toundertake thisloss reductionmeasure in the34 countriesmost exposedto severe winddamage; all ofthem exhibit aBCR41. Doingso could save65,700 livesover the next50 years.

Expected reductionsin damage toinfrastructure,property and avoidedfatalities

Other social,environmentalbenefits (dataavailability, scope ofstudy)

10-, 25-, 50-yrs 5, 12% (B/C) 2.2–6.07for human life$40k-6 M

Structural Forward-looking

a Three separate CBAs within Kunreuther and Michel-Kerjan [54] are reported on separate rows.b Two separate CBAs reported in Mechler [61] are shown on separate rows.c Results from original Kull [52] study are reported and expanded upon in Kull [53]d Results are reported only from the primary MMC [63] study. Other studies are noted as supporting studies, but not explicitly detailed here.e The range of values listed for BCRs are for minimum and maximum hazard occurrence scenarios, respectively.

C.M.Shreve,I.K

elman

/International

Journalof

Disaster

Risk

Reduction

10(2014)

213–235

226

1 Newhall [65] not shown in Table 1; BCR values not reported.

C.M. Shreve, I. Kelman / International Journal of Disaster Risk Reduction 10 (2014) 213–235 227

(e.g. [11,20]) or references listing other studies withoutanalysis (e.g. [2,23]) are not considered here. There are threekey metrics of economic efficiency within CBA: the benefit-to-cost ratio (BCR) or cost-benefit ratio (CBR), the internal rate ofreturn (IRR) and net present value (NPV), which in mostcircumstances are equivalent [53]. Here we have chosen theBCR, as it is commonly used to communicate with decisionmakers. This leaves out studies which do not report a BCR, buthelps to scope this paper. Certainly, no study can be compre-hensive but this paper compiles a large range of publications.

3. Results

Table 1 details main CBA parameters for the studiesreviewed including primary DRR activities, costs, benefitsand general framing. When noted by authors of thestudies, general categories of items not valued, but rele-vant to the CBA, are listed. The majority of studies (N¼22)were conducted at the community scale with fewernational scale studies (N¼7), the latter of which tendedto be for meteorological services (e.g. [63] and relatedstudies; [66]). Identified or intended benefactors of theDRR activities were largely residents in regions wherehazards are common (e.g. the general public (locally(N¼19), nationally (N¼6)), however some studies identi-fied benefactors by livelihood (e.g. farmers/agro-pastoral-ists (N¼2), fisherman/fishing communities (N¼1)), withone study identifying students as primary benefactors(N¼1).

The framing was relatively evenly split between ‘back-ward-looking’ (N¼10) and ‘forward-looking’ (N¼8) withsome applying both approaches (N¼3). ‘Forward-looking’studies are more difficult in the sense that they require anunderstanding of future risk, which is especially challen-ging with the uncertainty of climate change and climatechange modeling [82,52,53]. Other studies were eitherincluded as an element of a broader project, or as a partof a feasibility or impact study (N¼7).

3.1. Discount rates

The majority of studies used a single discount rate of10–12% (N¼16) with the minority applying a discount rateless than 5% (N¼3). Some studies investigated a range ofdiscount rates with two studies investigating a discountrate of 0–20% [52,53] and the rest investigating a smallerrage, e.g. 0–10% (N¼1), 5–9% (N¼2), 5–12% (N¼1), 0–15%(N¼1). Venton [83] argue that a very low or zero discountrate should be applied for environmental projects, asprotecting the environment for future generations shouldhave as much value as protecting the environment today.The higher the discount rate, the stronger the preference isfor present benefactors and the greater the burdenbecomes for future generations [83,53]. A high discountrate of 10–12% is standard practice for many developmentprojects, thus assuming that future generations will bebetter off and better able to cope with hazards [53].However, examining the full sensitivity over the full rangeof 0–20% helps to better understand the implications ofthe chosen rate [52].

3.2. Maximum BCR values reported

Maximum BCR values from the case studies evaluatedshowed a broad range across hazard type and geography.Table 2 provides a summary highlighting maximum BCRfindings. The highest BCR, 1800, was reported for droughtrisk reduction measures for an irrigation program support-ing communal gardens in the Sudan [48]. Across all hazardtypes, BCRs for DRR ranged from 3 to 15 in the regionswith studies. Southeastern Asia (e.g. Indonesia, Philip-pines, Vietnam) has the most widely reported BCRs acrosshazard type, including severe storms, drought, flood,earthquake, and volcanic hazards.

3.3. Results by hazard type

Flood risk reduction was the most commonly reportedBCR (N¼15), with an average value of 60 for 35 developingcountries for constructing a one-meter wall around houses,and 14.5 for elevating houses, in flood prone regions [54].The highest BCR values pertaining to flood risk reductionwere reported for mangrove forestation projects in Vietnam.The ‘Community-based Mangrove Reforestation and DisasterPreparedness Programme’ reported a BCR range of 3–68(excluding ecological benefits) and 28–104 (including ecolo-gical benefits) [38].

DRR effectiveness with regards to volcanic hazards isthe least reported in the case studies evaluated, with thesole record being for the Philippines. Newhall [65] 1 report‘the monitoring and response costs at US$56.5 millionwhile the amount of property damage averted as a resultof the monitoring and response is estimated at a minimumUS$500 million not including over 5000 lives saved.’

3.4. Types of DRR activities

Most studies had elements of both ‘structural’ (e.g.measures such as installing dykes, or levees) and ‘non-structural’ (e.g. measures such as developing an evacua-tion plan, training, and establishing community funds)DRR activities. The majority of studies reported difficultywith valuing certain components of non-structural activ-ities. ‘Non-structural’ activities often require valuing socialand environmental aspects that do not have a marketvalue (e.g. sense of security, peace of mind and avoidedproperty damage). Similarly, though direct costs are some-what easier to estimate for structural measures (e.g. cost ofconstruction materials, maintenance and labor), indirectcosts and benefits are rarely included.

3.5. Categories of items that were valued

General categories of items valued were developed toallow for basic comparisons across studies. ‘Agricultural’includes seeds, crop productivity and area of land used foragriculture. ‘Early warning system/meteorological services’refers to early warning systems and also general meteor-ological services. ‘Ecosystem services’ refers to items

Table 2Summary of maximum BCR values reported.

Authors Country Hazards BCR max

Venton [82]S Malawi Drought 24.0Khogali and Zewdu [52] Sudan Drought 1800.0Holland [33]S Fiji Flood 7.3Mechler [62]CC India Drought 3.5Khan [47]S Nepal Earthquake 2.0Kunreuther and Michel-Kerjan [54]S World (35) Flood 60.0

World (35) Wind damage 6.07World (35) Earthquake 5.1

Holland [33]S Fiji Flood 7.3EWASE [16]S Germany Flood 9.0Holub and Fuchs [34] Austria Flood; mass movements 1.7Mechler [61]S Peru Flood 3.8

Indonesia Flood 2.5Burton and Venton [5]S Philippines Flood 31.0White and Rorick [84] Nepal Flood 3.5Nepal Red Cross [64]S Nepal Flood 18.6Heidari [28] Iran Flood 1.3Khan [46] Pakistan Flood 25.0Kull [52,53]S,CC India Flood 2.5Kull [52,53]S,CC Pakistan Flood 25.0IFRC [39] Bangladesh Cyclone; flood 4.9Venton and Venton [81]S India Flood 57.8Dedeurwaerdere [12] Philippines Flood; lahar 30.0MMC [63] (and supporting studies) USA All 5.1; overall average 4Guocai and Wang [24] China All 4.0

(S) denotes sensitivity analysis was conducted and (CC) denotes climate change modeling was incorporated.


valued pertaining to recreational, biodiversity- andwatershed-services and ecosystem goods or products.‘Educational’ refers to time spent training or disruptionto education. ‘Emergency aid’ refers to all aspects, e.g.perishable and non-perishable goods, labor, and transportof goods. ‘Physical assets and maintenance’ refers toconstruction, building materials, energy and maintenancecosts. ‘Other’ refers to qualitative items not commonlyvalued, such as the value of human life. These categoriesare subjective and are meant only to frame a discussion ofwhat generally is and is not valued.

A majority of the studies valued physical assets andmaintenance (N¼25). As typically these items have widelyaccepted market values, this is the easiest category ofitems to value. Livelihood disruption was another commoncategory (N¼7), likely because a majority of the studieswere community-based and livelihood disruption can be amajor factor impacting a community during and after ahazard. More direct costs, for example loss of wages (e.g.markets temporarily closed, buildings/infrastructure tem-porarily damaged), are more easily estimated, but alsousually require a qualitative survey or discussion withrelevant experts to collect this information.

Indirect costs from livelihood disruption, such as men-tal and physical health costs, are noted by many studies asbeing underrepresented. Early warning and meteorologi-cal services studies (N¼3), which looked at the nationalscale, focused on specific sectors (e.g. [66] focused onagricultural and aviation, among other categories).Community-based studies also commonly looked at sectorspecific losses, for example, for agriculture (N¼5). Allstudies reviewed reported that actual benefits were under-estimated due to either limited data sources, inability to

assign a monetary value to social or environmental goods,or some combination of the two. Whitehead and Rose [85]are comparatively unique in their valuing of ecosystemservices (N¼1); many studies refer to ecosystem benefitsfrom DRR activities, but few quantify these benefits.

Human life was valued in 3 studies (e.g. separate DRRactivities reported in Kunreuther and Michel-Kerjan [54]).Other studies, for example Khan [47] did not value humanlife, but if they had, would likely have reported muchhigher BCRs. In the Kunreuther and Michel-Kerjan [54]study, the benefit of retrofitting schools in the mostseismically active countries to better safety standards doesnot exceed the cost for most countries until the value ofhuman life is added.

Finally, one study reported on emergency aid andincluded funds from non-governmental organizations(NGOs) and other funding bodies, which were grouped inthe ‘other’ category (e.g. [33]). A minority of studiesevaluated specific health costs, e.g. reported cases andassociated costs of specific diseases like malaria, or generalcosts from diarrhea outbreaks, commonly associated withwater contamination during floods (e.g. [46]), which werealso grouped in the ‘other’ category. Guocai and Wang [24]were also grouped in the ‘other’ category, as the study tookan economic approach surveying respondents’ willingnessto pay.

3.6. Individual case studies

Comparing individual case studies reveals severaltrends in the BCR data: the wide gaps in geographiccoverage and the prevalence of studies evaluating physicaland economic vulnerabilities, as opposed to social and


environmental vulnerabilities Mechler [61] addresses ele-ments of environmental vulnerability in considering‘fragility’, e.g. degree of damage as a function of hazardintensity for the environmental region impacted by thehazard, as well as direct and indirect economic impacts offlooding. Mechler [62] conduct a CBA in Indonesia includ-ing both qualitative and quantitative methods, providingBCR values for all four vulnerability categories. Otherexamples incorporating both qualitative and quantitativeanalyses of social vulnerability into the CBA include: TheNepal Red Cross [64] evaluation of the benefits of First Aidtraining; and Khan [46] and Kull [53] presentation of‘people-centered’ resilience-driven strategies, which eval-uate interventions at the individual scale.

3.7. Robustness and complexity of models: sensitivityanalyses and consideration of climate change impacts

Sensitivity analyses were reported in several studies(e.g. [5,16,81], 2009; [33,61], 2008; [47]; [52,53]; Kun-reuther and Michel-Kerjan [54]). Climate change modelingor scenarios were rarely incorporated into the CBA studiesevaluated. Studies incorporating both a sensitivity analysisand consideration of the impacts of climate change in theCBA were limited (e.g. [52,53,47]).

3.8. Scale

National scale studies focused on evaluating the DRRbenefits of: (1) improved weather forecasting systems(e.g. [23,24,56,58,69,89,90])2 covering a very limited numberof countries including the USA, Croatia, Belarus, Georgia,Kazakhstan, Nepal, and Samoa, (2) the costs of implementingstructural DRR measures, such as elevating houses or con-structing walls around houses to mitigate against earthquakesand floods (e.g. [54]), or (3) efficiency of DRR programs in theUSA (e.g. [17,63,20,72,22]) and Nepal [64]. National scalestudies were absent for the majority of Asia, Australia andSouth America.

Ecosystem restoration approaches were frequentlyevaluated at the sub-national or regional scale, for examplemangrove forestation in Vietnam [38] and river basinimprovements in Germany [16], India [52] and Pakistan[46]. Ecosystem restoration of floodplain or relocation out ofthe floodplain was evaluated on a hypothetical basis.

4. Discussion

4.1. By hazard type

The case studies are dominated by certain hazards, such asfloods, droughts, and earthquakes. Other environmentalhazards such as wildfires, tornados, extreme temperatures,and volcanoes are comparatively absent, even though deathsand damage related to natural hazards are not alwaysdominated by the hazards for which BCR studies exist,

2 Studies not reporting BCR values were not included in Table 1 withthe exception of NOAA [66], used to illustrate the point that many CBAsdo not report BCRs and instead report money saved or using anothermetric.

depending on the location (e.g. [41,75]). Floods and droughtsare likely highlighted due to the frequency with which theyoccur at a large damage scale, but that argument does nothold for earthquakes. Part of the bias is likely to be the ease ofcalculating costs and benefits of measures to reduce risks tofloods, droughts, and earthquakes. Studies examining CBA forother hazard types such as technological hazards, epidemicsand pandemics, both human and zoonotic diseases (e.g.diseases that transfer from animals to human) that may ormay not occur in relation to a natural hazard, are absent fromthe studies reviewed here.

Studies such as Newhall [65]2 demonstrate the feasi-bility of making those calculations for other hazards,suggesting a relatively easy way to expand the DRR BCRsfor DRR studies. Part of the bias is also likely to be thepopularity of CBA with engineers who often dominateflood and earthquake risk reduction initiatives—but whoalso dominate tornado risk reduction measures but notnecessarily drought risk reduction measures. As such, thebias might simply be inertia, in that the hazards dominat-ing the literature for DRR BCRs then inspire others topursue similar work.

4.2. By geographic region

Longitudinal studies comparing benefits of specific DRRstrategies or approaches across countries or locations withsimilarities are largely absent from the literature. Forinstance, even though the number of case studies forSouthern and Southeastern Asia is high in comparison toother regions, there is a lack of longitudinal comparisonsfor these countries. Some regional meta-studies exist, forexample, IADB [35] evaluate DRR in the LAC region;however, this study does not present an original CBA,thereby emulating the trend in available regional studies.

Nonetheless, the comparative lack of studies from LatinAmerica in this paper could be because available docu-ments, namely peer-reviewed papers and policy docu-ments, are dominated by English, especially for onlinesources. Personal discussions with members of La Red deEstudios Sociales en Prevención de Desastres (LA RED,The Network for Social Studies on Disaster Prevention)suggested that extensive material is available in Spanishfor Latin America, but it is not as formalized or systemizedas the English-language literature and does not alwaysprovide data of the form sought here.

4.3. Non-BCR approaches

The NOAA [66] study (Table 1) highlights an example ofa CBA following traditional techniques, but not reporting aBCR value. Results are instead reported as money saved.There are likely many other studies that report findings insimilar terms, however, we did not examine these studieshere, as it precludes comparison of BCR values. Addition-ally, there are alternate approaches to evaluating the costsand benefits of a project, which are not discussed due tothe scope of this paper. One such example is Cutter et al[8] which utilized country-level socioeconomic and demo-graphic data to generate an index of social vulnerability to


environmental hazards (e.g. the Social VulnerabilityIndex, SoVI).

4.4. Longevity of DRR benefits

There was limited evaluation of how long the DRRbenefits last, some studies examined only one or a smallrange of discount rates, and few studies included discus-sion of costs and benefits changing over time. Monitoringand evaluation of DRR tends to be linked to donors’ projectcycles, focusing on outputs of disaster planning versus theimpact such as the extent to which lives and assets arebetter protected as a result of DRR improvements [36].Frequency of hazard events and the potential for cascadingor ‘ripple effects’ are rarely considered in CBAs with oneexception being the MMC [63] studies, which do considerripple effects.

4.5. Less traditional DRR approaches: ecosystem-based DRR

Ecosystem restoration—for example, forestation withmangroves or rain forests, other forms of biomimicry,floodplain orchards, or other agroforestry techniques—can all be considered both DRR and sustainable conserva-tion strategies, which would also link to local livelihoods ifimplemented properly. The contemporary term is ‘ecosys-tem-based DRR’ [70], also cited in the literature as ‘nat-ural- or ecological-infrastructure’ which is gainingpopularity for implementation but for which costings arerarely available [14,25,49,42,77].

In the context of DRR, it would be advantageous tohave a more substantive understanding of CBA forecosystem-based DRR. Mangroves, while distributedacross 118 countries, have the highest concentration in15 countries in the tropics and subtropics [21]. Man-groves are said to offer physical sea defenses, includingabsorbing and dampening wave action, slowing erosionrates, and fostering biodiversity and sequestering carbon.Indonesia has roughly 22% of the global total of man-groves; Brazil and Australia have about 7% each; andBangladesh and India have approximately 2–3% each [21].However, CBA for using mangroves for DRR were onlyreported in two separate case studies for Vietnam(UNIDSR, 2002; The Red Cross, 2008) and their (and coralreefs’) effectiveness in mitigating tsunami and surgedamage is not straightforward (e.g. [9]).

Agroforestry techniques such as using floodplain orch-ards, polyculturing of annuals and perennials, or rain-forestation support, as is common practice for manyindigenous groups in the Amazon [6], may not be widelyrecognized DRR techniques, since they are infrequentlycited in the DRR literature. These techniques promote DRRthrough ecological conservation, e.g. reducing soil erosionand maintaining the hydrologic cycle, which mitigatephysical risk from natural disasters, as well as enablinglivelihoods and the use of traditional knowledge. Similarlyto mangrove planting, the costs and benefits of sustainableagriculture techniques as DRR strategies have limitedcoverage in the literature.

The sole reference found here was Dedeurwaerdere [12],who analyzed the potential cost-benefit of rainforestation in

the Philippines, reporting a BCR of 30 for a 15-year, 1000hectares (10 km2) project. In comparison, IFRC [37] evaluatedartificial structural measures, e.g. seawall and dykes, in thePhilippines, reporting BCR values of 4.9 and 0.67, respectively.That suggests the big gains feasible through ecosystem-basedDRR compared to artificial structures.

CBAs for DRR involving environmental componentscould benefit from techniques for quantification, valuingand marketing of ecosystem services. For example, inter-nationally agreed upon methods and standards exist forthe quantification and monitoring of forest carbon andcarbon offsets are currently marketed. No similar marketsexist for watershed- and biodiversity-services, though theconcept of ‘biodiversity offsets’ has been proposed in someareas such as England. Forest carbon valuation methodscould be used to assign a value to ecosystem services inphysical ‘risk-based’ hazard models where appropriate, asgenerally these services are not currently valued in impactassessments. However, ethical concerns regarding offset-ting effects on society and nature and irreversibilityremain to be addressed. Another consideration is thatforests and other ecosystems have disaster risk reductionbenefits that extend beyond carbon uptake. In fact, thispoint is a major concern about using quantifiable ecosys-tem services in conjunction with CBA: that the focus mightend up on a small number of services without beingcomprehensive.

4.6. Additional considerations

Relocation of people outside of hazard zones is frequentlyconsidered as a DRR measure, raising significant ethical andjustice questions regarding who makes the decisions, whopays for the decisions, and who is affected by the decisions.No case studies were found which presented backward-looking BCR for relocation, despite a large literature onpopulation movement for DRR and after a disaster.

Overall, the literature displays no consensus regardinghow CBA analyses should be conducted, what base vari-ables to include, or how to deal with limitations of CBA asdiscussed earlier. For example, should CBA for DRR beconducted by area, by population, hazard type, vulner-ability type, other variables, or a combination? Should thebenefits of education and training, which are mostcommonly not calculated, with a few exceptions (e.g.[64]), be a mandatory variable for inclusion and assess-ment? Consequently, comparing CBAs might have limitedvalidity due to them using different baselines and/ormethodologies.

Sometimes CBA is not independent of the DRR mea-sure itself. For example, structural flood defenses areeasily costed, as are the property and possessions (andpotentially lives) which are ‘protected’. But then the DRRmeasure itself influences the situation, thereby affectingCBA. In the case of structural flood defenses, people gain afalse sense of security due to the visibility and hardness ofthe measures, leading them to build and settle in areas‘protected’ by the flood defenses without taking ade-quate, further DRR measures. That is, the presence of thestructural measures leads to reliance on them and hence


an increase in the property, possessions, and people whoare vulnerable in cases of defense failure [15,19,79].

Kull [53] extend this discussion by pointing out thatcertain DRR options may generate ‘disbenefits’ or negativeexternalities. They cite the example of embankmentsprotecting an area from a flood, but simultaneouslyincreasing the risk of water logging, which is associatedwith increases in vector-borne diseases and decreases incrop productivity. Many studies discuss why certain DRRmeasures might not be cost-effective if the BCR is close toor less than one, but few evaluate disbenefits or potentiallynegative spillover effects.

4.7. Can CBAs highlight vulnerability?

Vulnerabilities, rather than hazards, are the root causeof disasters [29,30,59,86–88]. As the literature shows,vulnerabilities are not caused by nature or the environ-mental hazards, but instead are social constructions[32,67].

Overall, the CBA studies tend to generalize vulner-ability into four broad categories: economic (financialcapacity to return to a previous path after a disaster);environmental (a function of factors such as land andwater use, biodiversity and ecosystem stability); physical(related to susceptibility of damage to engineered struc-tures such as houses, damns and roads; populationgrowth); and social (ability to cope with disaster at theindividual level as well as capacity of institutions to copeand respond) [61]. While all four categories are recog-nized as important, social and environmental impacts aremore qualitative in nature and therefore the focus of CBAfor DRR tends to be on the quantitative economic andphysical impacts.

A common approach in many government guidelines isto utilize Multi-Criteria Analysis (MCA) to address thequalitative variables, such as social and environmentalbenefits, of a project as a subset within the CBA [1]. Thatis, MCA is used to address the qualitative costs andbenefits and CBA is used to address the quantitative costsand benefits. MCA utilizes expert opinion—such as demo-cratic voting, a panel of experts, a consensus model, orfocus groups—to select the criteria and the rating optionsfor the model. MCA's flexibility allows for a greaterrange of awareness and involvement across scales; forexample, joining views of an individual household or apanel of international experts—but adds complexity andsubjectivity.

MCA could play a significant role in highlighting thelonger-term implications of DRR. Brouwer and van Ek's [4]integrated CBA and MCA approach to flood control policyin the Netherlands found that, while structural measureswere more cost-effective in the short term according tothe CBA, floodplain restoration can be justified using boththe CBA and MCA considering socio-economic and ecolo-gical impacts in the longer-term. That outcome was notvisible utilizing CBA alone. MCA may be more efficient athighlighting social and environmental vulnerabilities andthus benefits than CBA alone.

4.8. Recommendations of minimum criteria to improve DRRCBA

Currently there is no consensus on the minimumcriteria necessary for conducting a comprehensive CBAfor DRR. For instance, there is no standard or systematicapproach detailing what variables need to be assessed torepresent vulnerability, disaster consequences, or even theappropriate spatial and temporal scales for determiningCBA, vulnerability, or disaster consequences.

Vulnerability is not homogenous, nor are the benefitsgained from DRR. Most DRR CBA studies either focus onpoor or marginalized groups, as these are commonly themost vulnerable to disaster, or alternatively, considervulnerability to be rather homogenous, e.g. broad scaleprograms evaluating the effectiveness of weather forecastsor government subsidies. While it is unrealistic to thinkthat vulnerability can be comprehensively assessed (e.g.[59,86,87,88]), CBAs could be improved by a more sys-tematic approach that better defines the context in whichvulnerability is assessed so that the CBA is then contex-tualized appropriately and can be interpreted and com-pared within that contextualization. One consequencewould be knowing what lessons are and are not transfer-able amongst different contexts.

Another challenge within the standardization of CBAfor DRR is which consequences to consider in the calcula-tions. In forward- and backward-looking DRR CBAs, it iscommon to consider obvious and immediate disasterconsequences, such as physical damage to infrastructure,loss of life, injuries, and systems failures. Further conse-quences, often appearing after the initial hazard hasdissipated, are less frequently considered. Examples arepsychological health impacts, continued water logging orsalinity of crops, business interruption, bankruptcies, andlong-term migration.

Including climate change modeling results does notnecessarily enhance a CBA for weather- and climate-related hazards. Variability in the accuracy and precisionof climate data, difficulty associated with projecting andpredicting hazard occurrence, and challenges in incorpor-ating future social behavior and policies, contribute to theuncertainty of future climate impacts. Using the models todevelop scenarios for climate evolution and consequentweather extremes, especially when down-scaled to regio-nal or local levels, can help to depict a more comprehen-sive portrait of the hazard side of the expected risk overthe lifetime of the suggested benefits from a DRR inter-vention, thereby further highlighting the benefits of DRRmeasures across multiple scenarios.

That suggests a further limitation in comparing DRRCBAs, notably across hazards. For well-studied fault lines,such as the San Andreas fault in California, future prob-abilities of the hazard occurring are available [76,7] mean-ing that, given the knowledge of building codes andconstruction practices, the benefits of additional DRRmeasures are calculable [71,73]. For changes in floodregimes, as a result of climate change as well as infra-structure development, understanding the hazard andvulnerability changes is much more challenging withlarger uncertainties [78,43]. DRR CBAs might have


different levels of usefulness depending on the hazard anddepending on the hazard drivers, such as climate change,which are considered for analyzing CBAs in forward-looking studies.

Additionally, spatial and temporal scales of the CBAcalculation can impact the validity of the assessment ofDRR benefits. The issue of long-term disaster impacts, andhence potential benefits, from a DRR intervention isdiscussed above. Regarding spatial scale, CBA studies thatfail to consider wider impacts within a wider system arenot necessarily truly representing the costs or benefits.Levees affect the hydrological regime both upstream anddownstream and exacerbate flooding in other places [79].In comparison, earthquake-resistant technologies, despitetheir heavy reliance on structural approaches, save lives[74] with no foreseen long-term consequences. The basicexamples above illustrate the complexities related to scaleand vulnerability and the importance of considering scalewhen evaluating the benefits of DRR.

As was emphasized by Deaton [10], disaster mitiga-tion must be informed by a strong understanding of thepublic and private sector decision process. Considerationmust be given to how these decisions will impact cost-and benefits- over the long-term and across scale. Forexample, crop insurance schemes must not only con-sider what perils, crops, and amount of indemnificationto cover, but also consider farmer understanding of theprogram and capacity required for farmers involved tomake claims.

CBAs should not just mirror the conceptual process ofmitigation planning, but be used as a tool for improvementat various stages within this process. CBAs should be anopportunity to promote investment in DRR and evaluatesuccesses and failures of pilot programs. At the same time,CBAs are not a panacea and need to clearly outline whatcan and cannot be valued for a given project.

Many examples were given where DRR practitionershave overcome noted limitations, e.g. Khan [47] evaluatedthe CBA of straw-bale for its efficiency as a buildingmaterial and the potential for positive environmentalbenefits in Nepal, such as reduced greenhouse gas (GHG)emissions in comparison to traditional brick materials. Thelargest benefit of using straw-bale materials in housing isits resistance to earthquakes that are common to theregion. However, analyzing the CBA of straw-bale as asustainable building material, instead of assessing its valuefor earthquake mitigation, sidesteps the issue of assigninga value to human life.

The majority of CBAs presented here have deliberatelyavoided valuing human life. The reason is the ethicalchallenges of selecting a monetary value for human life[60], which many find inappropriate in principle. Oneinteresting outcome of a large number of the studies onDRR CBAs is that, even without considering the value oflife and other intangibles, high BCRs emerge. This is notalways the case, for example, Kunreuther and Michel-Kerjan's [54] CBA assessing costs to retrofit schools inseismically active countries did not exceed a BCR value ofone for many countries until a value was assigned tohuman life. Thus, assigning a value to human life in thatexample may have benefits.

We recommend greater consideration of the contextand methods used to assess vulnerability, disaster con-sequences, and spatial and temporal scales as areas thatneed further investigation and standardization to improveupon current DRR CBAs.

5. Conclusions

This study reviewed individual CBA case studies report-ing BCRs spanning different geographies, hazard types,and vulnerabilities. Many results emerged displaying solidevidence to support the economic effectiveness of DRR,but several key limitations were identified, including a lackof: sensitivity analyses of the CBA, meta-analyses whichcritique the literature, consideration of potential impactsof climate change, evaluation of the duration of benefits,broader consideration of the process of vulnerability, andpotential disbenefits of DRR measures. To represent thepotential benefits of DRR more comprehensively to deci-sion makers, these concerns will need to be addressed.

Yet the studies also lucidly demonstrate the importanceof context for each BCR result. It is not clear that averagingBCRs across case studies produces a useable result forpolicy or decision makers, because the circumstances ofthe studies tend to be quite different—particularly withrespect to vulnerability. The BCR technique generally haslimited applicability for factoring in vulnerability.The contextual aspects of each case study tend to be thevulnerabilities.

Part of understanding and incorporating context is theinfluence of culture on hazard, vulnerability, risk anddisaster [31,51]. When determining costs and benefits,the values assigned can differ depending on who is asked,with different perspectives assigning different values forproperty, land, and infrastructure. Additionally, the metricsused for costs and benefits have been absolute, specificvalues, such as $3 million or €5,000. Vulnerability theories(e.g. [59]) also discuss the need for proportional metrics,such as stating that 12% of assets would be lost or thatbenefits accrued would be 135% of current value. Report-ing proportional vulnerability and proportional gains forCBAs would avoid biasing measures towards helping thosewho are affluent, and who therefore stand to lose a lot inabsolute measures, compared to those who have little tobegin with, so even a small absolute loss can be most oftheir assets.

Several studies demonstrate that these difficulties canbe addressed more robustly to some degree, as long as thecontext is retained. For example, using shared learningdialogues (SLDs), a participatory and multi-stakeholderapproach to assessing vulnerability, (e.g. [46]) helps tounderstand the origin of the numbers leading to the BCR.Evaluating training benefits [64] is another approachneeded across more case studies.

As such, comparing locations, hazards, or scales mightnot yield results which are meaningful for decision-making. Instead, to determine financially whether or nota DRR measure or process should be implemented, calcu-lations need to be made for that specific case study,potentially employing MCA or SLDs during the planningstages to guide longer-term social and ecological goals.


Rather than simply reporting a single ratio, these calcula-tions should also consider how the DRR measure orprocess might affect the costs and benefits, which valuesare not included in the calculations, and the contextuali-ties for that case study. It seems that disaster mitigationcan indeed save money in numerous circumstances, but‘How much money can we calculate will be saved?’ is notthe only question.

Acknowledgments

The authors would like to thank the following collea-gues for their contributions to the field and supportivecomments in development of this work: Bob Alexander,Steve Bender, Charlotte Benson, David Crichton, KateHawley, Terry Jeggle, Fawad Khan, Daniel Kull, HowardKunreuther, James Lewis, Reinhard Mechler, ErwannMichel-Kerjan, Marcus Moench, Chris Newhall, CharlesSetchell, John Twigg, Courtney Cabot Venton, TriciaWachtendorf, and Siân Pooley. This paper is based onKelman and Shreve [44] which is an informal technicaldocument compiling examples of CBA for DRR. Kelman[45] provides general comments on CBA for DRR in atechnical report for the World Bank.

References

[1] Argyrous, G. Cost-benefit analysis and multi-criteria analysis:competing or complimentary approaches?” School of Socialand International Studies, The Australia and New Zealand Schoolof Government; 2010. ⟨http://www.anzsog.edu.au/magma/media/upload/ckeditor/files/Special%20Events/Argyrous_Costbenefit_DarwinJune2010_Presentation.pdf⟩ (June 24, 2013).

[2] Benson C, Twigg J. Measuring mitigation methodologies for asses-sing natural hazard risks and the net benefits of mitigation – ascoping study. IFRC (International Federation of Red Cross and RedCrescent Societies) and the ProVention Consortium, Geneva; 2004.

[3] Blevins C. Utilizing geospatial technology for disaster risk reduction.USAID Office of Foreign Disaster Risk Reduction. ESRI Federal UsersConference, Washington, D.C.; 2012.

[4] Brouwer, Roy, Remco Van Ek. Integrated ecological, economicand social impact assessment of alternative flood control policies inthe Netherlands. Ecol Econ 2004;50.1:1–21, http://dx.doi.org/10.1016/j.ecolecon.2004.01.020.

[5] Burton C, Venton CC. Case study of the Philippines national redcross: community based disaster risk management programming.Geneva, Switzerland: IFRC (International Federation of Red Crossand Red Crescent Societies); 2009.

[6] Butler R. Sustainable agriculture in rainforests. Mongabay 2012 (June23, 2013)⟨http://rainforests.mongabay.com/1002.htm⟩.

[7] Catchings RD, Rymer MJ, Goldman MR, Prentice CS, Sickler RR. Fine-scale delineation of the location of and relative ground shakingwithin the San Andreas Fault zone at San Andreas Lake, San MateoCounty, California: U.S. Geological Survey Open-File Report 2013–1041; 2013. p.53 [Available at http://pubs.usgs.gov/of/2013/1041/9.].

[8] Cutter S, et al. Social vulnerability to environmental hazards. Soc SciQ 2003;84(2):241–61.

[9] Dahdouh-Guebas F, Koedam N. Coastal vegetation and the Asiantsunami. Science 2006;311:37.

[10] Deaton B. The economics of disaster mitigation. In: Proceedings ofthe International Conference on Disaster Mitigation, Virginia Poly-technic Institute and State University, Ocho Rios, Jamaica; 1984.

[11] de Loë R, Wojtanowski D. Associated benefits and costs of theCanadian flood damage reduction program. Appl Geogr 2001;21:1–21.

[12] Dedeurwaerdere A. Cost-benefit analysis for natural disaster man-agement – a case-study in the Philippines. Brussels: CRED; 1998.

[13] Dilley M, Heyman BN. ENSO and disaster: droughts, floods and Elnino southern oscillation warm events. Disasters 1995;19(3):181–93.

[14] Dudley N, Stolton S, Belokurov A, Krueger L, Lopoukhine N, MacK-innon K, Sandwith T, Sekhran N. editors. Natural Solutions: Pro-tected areas helping people cope with climate change, IUCN- WCPA,TNC, UNDP, WCS, The World Bank and WWF, Gland, Switzerland,Washington DC and New York, USA; 2010.

[15] Etkin D. Risk transference and related trends: driving forces towardsmore mega-disasters. Environ Hazards 1999;1:69–75.

[16] EWASE. CRUE Research report No. I-5: Effectiveness and Efficiency ofEarly Warning Systems For Flash Floods (EWASE). First CRUE ERA-Net Common Call Effectiveness and Efficiency of Non-structuralFlood Risk Management Measures; 2008.

[17] FEMA. Report on Costs and Benefits of Natural Hazard Mitigation.FEMA (Federal Emergency Management Agency), Washington, D.C.,and developed under the Hazard Mitigation Technical AssistanceProgram (HMTAP) Contract Number 132 with Woodward-ClydeFederal Services, Gaithersburg, Maryland; 1997.

[18] FEMA. Using Benefit-Cost Review in Mitigation Planning. How-ToGuide Number Five. FEMA 386-5, Washington, DC; 2007.

[19] Fordham M. Participatory planning for flood mitigation: models andapproaches. Aust J Emerg Manag 1999;13(4):27–34.

[20] Ganderton PT. ‘Benefit-Cost Analysis’ of disaster mitigation: applica-tion as a policy and decision-making tool. Mitig Adapt StrategiesGlob Change 2005;10:445–65.

[21] Giri C, Ochieng E, Tieszen LL, Zhu Z, Singh A, Loveland T, et al. Statusand distribution of mangrove forests of the world using earthobservation satellite data. Glob Ecol Biogeogr 2011;20(1):154–9, http://dx.doi.org/10.1111/j.1466-8238.2010.00584.x.

[22] Godschalk DR, Rose A, Mittler E, Porter K, West CT. Estimating thevalue of foresight: aggregate analysis of natural hazard mitigationbenefits and costs. J Environ Plan and Manag 2009;52(6):739–56, http://dx.doi.org/10.1080/09640560903083715.

[23] Gunasekera D. Measuring the economic value of meteorologicalinformation. WMO Bull 2003;52(4):366–73.

[24] Guocai Z, Wang H. Evaluating the benefits of meteorological servicesin China. WMO Bull 2003;52(4):383–7.

[25] Gupta AK, Nair SS. Applying EIA and SEA in disaster management.In: Renaud FG, Sudmeier-Rieux K, Estrella M, editors. The role ofecosystems in disaster risk reduction. Tokyo: United NationsUniversity Press; 2012.

[26] Hanley NC, Splash. CL. Cost benefit analysis and the environment.Northampton, Massachusetts, USA: MA Edward Elgar Publishing;1993.

[27] Hawley K, Moench M, Sabbag L. Understanding the economics offlood risk reduction: a preliminary analysis. Boulder, CO: Institutefor Social and Environmental Transition-International; 2012.

[28] Heidari A. Structural master plan of flood mitigation measures. NatHazards Earth Syst Sci 2009;9:61–75.

[29] Hewitt K, editor. Interpretations of calamity from the viewpoint ofhuman ecology. London, UK: Allen & Unwin; 1983.

[30] Hewitt K. Regions of risk: a geographical introduction to disasters.Essex: Longman; 1997.

[31] Catastrophe and culture: the anthropology of disaster. In: HoffmanSM, Oliver-Smith A, editors. Santa Fe/Oxford: School of AmericanResearch Press/James Currey; 2002.

[32] Hoffman SM, Oliver-Smith A. Anthropology and the angry earth: anoverview. In: Oliver-Smith Anthony, Hoffman Susanna M, editors.The angry earth. New York: Routledge; 1999. p. 1–16.

[33] Holland P. An economic analysis of flood warning in Navua, Fiji.European Union Development Fund (EU EDF) 8 – SOPAC ProjectReport 122, Reducing Vulnerability of Pacific ACP States, Fiji Tech-nical Report. SOPAC (Pacific Islands Applied Geosciences Commis-sion), Suva, Fiji; 2008.

[34] Holub M, Fuchs S. Benefits of local structural protection to mitigatetorrent-related hazards. In: Brebbia CA, Beritatos E, editors. RiskAnalysis VI, WIT Transactions on Information and CommunicationTechnologies, vol. 39. Southampton, U.K.: WIT Press; 2008.p. 401–11.

[35] IADB, IMF, OAS, and the World Bank. The economics of disastermitigation in the Caribbean quantifying the benefits and costs ofmitigating natural hazard losses. IADB (Inter-American Develop-ment Bank), IMF (International Monetary Fund), OAS (Organizationof American States), and the World Bank, Washington, D.C.; 2005August.

[36] IFRC. World Disasters Report 2002. Geneva: International Federationof Red Cross and Red Crescent Societies. Geneva, Switzerland; 2002.

[37] IFRC. Assessing Quality and Cost Benefit: A Philippines Case Study;2009.

[38] IFRC. Breaking the waves Impact analysis of coastal afforestation fordisaster risk reduction in Viet Nam. Geneva, Switzerland: IFRC

http://www.anzsog.edu.au/magma/media/upload/ckeditor/files/Special%20Events/Argyrous_Costbenefit_DarwinJune2010_Presentation.pdf



http://dx.doi.org/<mac_doitxt>10.1016/j.ecolecon.2004.01.020




http://refhub.elsevier.com/S2212-4209(14)00066-1/sbref0005




http://rainforests.mongabay.com/1002.htm

http://pubs.usgs.gov/of/2013/1041/9




















http://dx.doi.org/10.1111/j.1466-8238.2010.00584.x

http://dx.doi.org/10.1111/j.1466-8238.2010.00584.x

http://dx.doi.org/10.1111/j.1466-8238.2010.00584.x

http://dx.doi.org/10.1080/09640560903083715

http://dx.doi.org/10.1080/09640560903083715

http://dx.doi.org/10.1080/09640560903083715






































(International Federation of Red Cross and Red Crescent Societies);2011.

[39] IFRC. The long road to resilience: impact and cost-benefit analysis ofcommunity-based disaster risk reduction in Bangladesh. Geneva,Switzerland: IFRC (International Federation of Red Cross and RedCrescent Societies); 2012.

[40] Jonkman SN, Brinkhuis-Jak M, Kok M. Cost benefit analysis and flooddamage mitigation in the Netherlands. HERON 2004;49(1):95–111.

[41] Jonkman SN. Global perspectives of loss of human life caused byfloods. Nat Hazards 2005;34(2):151–75.

[42] Kazmierczak A, Carter J. Adaptation to climate change using greenand blue infrastructure, a database of case studies. Manchester:University of Manchester; 2010.

[43] Kelman I. Decision-making for flood-threatened properties. In:Begum S, Stive M, Hall J, editors. Flood risk management in europe:innovation in policy and practice, vol. 25 of the book series onadvances in natural and technological hazards research, Chapter 1.Dordrecht, the Netherlands: Springer; 2007. p. 3–19.

[44] Kelman I, CM Shreve editors. Disaster Mitigation Saves. Version 12,13 June 2013 (Version 1 was 30 October 2002). ⟨http://www.ilankelman.org/miscellany/MitigationSaves.doc⟩.

[45] Kelman I. Disaster mitigation is cost effective. Background notefor the World Development Report. CICERO, Norway; 2014.⟨http://siteresources.worldbank.org/EXTNWDR2013/Resources/8258024-1352909193861/8936935-1356011448215/8986901-1380568255405/WDR14_bp_Disaster_Mitigation_is_Cost_Effective_Kelman.pdf⟩.

[46] Khan, F., Mustafa, D., Kull, D., & The Risk to Resilience Study Team.(2008). Evaluating the costs and benefits of disaster risk reductionunder changing climatic conditions: A Pakistan case study (Risk toResilience Working Paper No. 7). M. Moench, E. Caspari, & A. Pokhrel(Eds.). Kathmandu, Nepal: Institute for Social and EnvironmentalTransition-Boulder, Institute for Social and Environmental Transi-tion-Nepal, & Provention Consortium.

[47] Khan F, Moench M, Reed SO, Dixit A, Shrestha S, Dixit K. Under-standing the costs and benefits of disaster risk reduction underchanging climate conditions: Case study results and underlyingprinciples. Bangkok, Thailand: Institute for Social and Environmen-tal Transition-International; 2012.

[48] Khogali H, Zewdu D. Impact and cost benefit analysis: a case studyof disaster risk reduction programming in Red Sea State Sudan.Khartoum, Sudan: Sudanese Red Crescent Society; 2009.

[49] Kousky C. Using natural capital to reduce disaster risk. J Nat ResourPolicy Res 2010;2(4):343–56.

[50] Kramer RA. Advantages and limitations of benefit-cost analysis forevaluating investments in natural disaster mitigation. In: Muna-singhe M, Clarke C, editors. Disaster prevention for sustainabledevelopment: economic and policy issues. Washington DC: WorldBank Publication; 1995. p. 61–76.

[51] Krüger F, Bankoff G, Cannon T, Schipper L. Cultures and disasters:understanding cultural framings in disaster risk reduction. Abing-don: Routledge; 2015.

[52] Kull D. Evaluating costs and benefits of flood reduction underchanging climatic conditions: case of the Rohini River Basin, India.From Risk to Resilience Working Paper No. 4. Moench M, Caspari E,Pokhrel A, editors. ISET, ISET-Nepal and ProVention, Kathmandu,Nepal; 2008. 32 p.

[53] Kull, Daniel, Reinhard Mechler, and Stefan Hochrainer‐Stigler. "Prob-abilistic cost‐benefit analysis of disaster risk management in adevelopment context."Disasters 37.3 (2013): 374-400. http://dx.doi.org/10.1111/disa.12002.

[54] Kunreuther H, Michel-Kerjan E. Challenge Paper: Natural Disasters.Policy Options for Reducing Losses from Natural Disasters: Allocat-ing $75 billion. Revised version for Copenhagen Consensus.” Centerfor Risk Management and Decision Processes, The Wharton School,University of Pennsylvania, Philadelphia, Pennsylvania, U.S.A.; 2012.

[55] La Trobe S, Venton P. Natural disaster risk reduction: the policy andpractice of selected institutional donors. London, U.K.: A TearfundResearch Project. Tearfund; 2003.

[56] Lazo JK, Chestnut LG. Economic value of current and improvedweather forecasts in the US household sector: report prepared forthe National Oceanic and Atmospheric Administration. Boulder, CO:Stratus Consulting Inc.; 2002.

[57] Lefton R. The debt deal and budget process threaten critical climate aidfunding. Center for American Progress; 2011. ⟨http://www.americanprogress.org/issues/green/news/2011/09/13/10346/the-debt-deal-and-budget-process-threaten-critical-climate-aid-funding/⟩ (June 23, 2013).

[58] Leviäkangas, P, Hautala, R, Räsänen, J, Öörni, R, Sonninen, S,Hekkanen, M, et al. 2007. Benefits of meteorological services in

Croatia. Espoo2007. VTT Tiedotteita – Research Notes 2420. 71 p. +app. 2 p.

[59] Lewis J. Development in disaster-prone places: studies of vulner-ability. London, UK: Intermediate Technology Publications; 1999.

[60] May WW. $s for lives: ethical considerations in the use of cost/benefit analysis by for-profit firms. Risk Anal 1982;2(1):35–46.

[61] Mechler R. Cost-benefit analysis of natural disaster risk manage-ment in developing countries. Working paper for sector project‘Disaster Risk Management in Development Cooperation’, GTZ,Berlin; 2005.

[62] Mechler, R., Hochrainer, S., Kull, D., Singh, P., Chopde, S., Wajih, S., &The Risk to Resilience Study Team. (2008). Uttar Pradesh droughtcost-benefit analysis, India (Risk to Resilience Working Paper No. 5).M. Moench, E. Caspari, & A. Pokhrel (Eds.). Kathmandu, Nepal:Institute for Social and Environmental Transition-Boulder, Institutefor Social and Environmental Transition-Nepal, & ProventionConsortium.

[63] MMC. Natural hazard mitigation saves: an independent study toassess the future savings frommitigation activities. Vol. 1 – Findings,Conclusions, and Recommendations. Vol. 2 – Study Documentation.Appendices. MMC (Multihazard Mitigation Council). National Insti-tute of Building Sciences, Washington, D.C.; 2005.

[64] Nepal Red Cross. Cost benefit analysis of a Nepal red cross societydisaster risk reduction programme. Kathmandu, Nepal: Nepal RedCross; 2008.

[65] Newhall, C., Hendley, J.W. II, and P.H. Stauffer. Reducing the Riskfrom Volcano Hazards: Benefits of Volcano Monitoring Far OutweighCosts — The Case of Mount Pinatubo.Pinatubo. U.S. GeologicalSurvey Fact Sheet 115-97, Vancouver, Washington, U.S.A.; 1997.

[66] National Oceanic and Atmospheric Administration (NOAA). AnInvestigation of the Economic and Social Value of Selected NOAAData and Products for Geostationary Operational EnvironmentalSatellites (GOES). A Report to NOAA's National Climatic Data Centerand National Environmental Satellite, Data and Information Service(NESDIS). CENTREC Consulting Group, LLC. Savoy, IL, 28 February2007.

[67] Oliver-Smith A, Hoffman SM. Theorizing disasters: nature, powerand culture (catastrophe and culture: the anthropology of disaster).Oliver-Smith A, editor. Santa Fe, School of American Research Press;2002.

[68] Murphy B. UN disaster risk reduction conference: good, but needs to gofurther. Oxfam; 2013. ⟨http://blogs.oxfam.org/en/blogs/13-05-24-un-disaster-risk-reduction-conference-good-needs-go-further⟩ (June 24, 2013).

[69] Perrels A. Social economic benefits of enhanced weather services inNepal. Helsinki, Finland: Finnish Meteorological Institute; 2011.

[70] Renaud FG, Sudmeier-Rieux K, Estrella Marisol, editors. The role ofecosystems in disaster risk reduction. Tokyo: United NationsUniversity Press; 2013.

[71] RMS (Risk Management Solutions, Inc.). What if the 1906 Earth-quake Strikes Again?. A San Francisco Bay Area Scenario, ExecutiveSummary; 1995. 7 p.

[72] Rose A. Benefit-cost analysis of FEMA hazard mitigation grants. NatHazards Rev 2007;8(4):97–111.

[73] Rowshandel, B., Reichle, M., Wills, C., Cao,T., Petersen,M., Branum,D.et al. (2005). “Estimation Estimation of Future Earthquake Losses-future earthquake losses in California,” webCalifornia. Web informa-tion paper, paper; 2005.

[74] Schulze WD, Brookshire DS, Hageman RK, Tschirhart J. Benefits andcosts of earthquake resistant buildings. Southern Economic Journal,1987:934–51.

[75] Simkin, Tom, Lee Siebert, and Russell Blong. "Volcano Fatalities–Lessonsfrom the Historical Record." Science 291.5502 (2001): 255-255. http://dx.doi.org/10.1126/science.291.5502.255.

[76] Smith‐Konter, Bridget R., David T. Sandwell, and Peter Shearer."Locking depths estimated from geodesy and seismology along theSan Andreas Fault System: Implications for seismic momentrelease." Journal of Geophysical Research: Solid Earth (1978–2012)116.B6 (2011). http://dx.doi.org/10.1029/2010JB008117.

[77] Sudmeier-Rieux, K., F. G. Renaud, and M. Jaboyedoff. "Opportunities,incentives and challenges to risk sensitive land use planning.Lessons from Nepal, Spain and Vietnam.".

[78] Szöllösi-Nagy A, Zevenbergen C. Urban flood management. TheNetherlands: A.A. Balkema Publishers; 2004.

[79] Tobin GA. The Levee love affair: a stormy relationship. Water ResourBull 1995;31(3):359–67.

[80] UN GA/11048 DP23. United Nations General Assembly; 2011. ⟨http://www.un.org/News/Press/docs/2011/ga11048.doc.htm⟩ (June 24,2013).



















http://www.ilankelman.org/miscellany/MitigationSaves.doc

http://www.ilankelman.org/miscellany/MitigationSaves.doc

http://siteresources.worldbank.org/EXTNWDR2013/Resources/8258024-1352909193861/8936935-1356011448215/8986901-1380568255405/WDR14_bp_Disaster_Mitigation_is_Cost_Effective_Kelman.pdf






















http://dx.doi.org/10.1111/disa.12002

http://dx.doi.org/10.1111/disa.12002








http://www.americanprogress.org/issues/green/news/2011/09/13/10346/the-debt-deal-and-budget-process-threaten-critical-climate-aid-funding/










http://blogs.oxfam.org/en/blogs/13-05-24-un-disaster-risk-reduction-conference-good-needs-go-further

http://blogs.oxfam.org/en/blogs/13-05-24-un-disaster-risk-reduction-conference-good-needs-go-further











http://dx.doi.org/10.1126/science.291.5502.255

http://dx.doi.org/10.1126/science.291.5502.255

http://dx.doi.org/10.1029/2010JB008117





http://www.un.org/News/Press/docs/2011/ga11048.doc.htm

http://www.un.org/News/Press/docs/2011/ga11048.doc.htm


[81] Venton CC, Venton P. Disaster preparedness programmes in India: acost benefit analysis.Network Paper Number 49, HumanitarianPractice Network. London, U.K.: Overseas Development Institute;2004.

[82] Venton. Cost benefit analysis for community based climate anddisaster risk management: synthesis report. London, U.K.: Tearfund;2010.

[83] Cabot Venton CC, Siedenburg J, Faleiro J, Khinmaung J. Investing InCommunities: the benefits and costs of building resilience for foodsecurity in Malawi. London, U.K.: Tearfund; 2010.

[84] White BA, Rorick MM. Cost-benefit analysis for community-baseddisaster risk reduction in Kailali, Nepal. Mercy Corps Nepal, Lalitpur,Nepal; 2010.

[85] Whitehead J, Rose A. Estimating environmental benefits of naturalhazard mitigation: results from a benefit-cost analysis of FEMAmitigation grants. Mitig Adapt Strategies Glob Change 2009;14:655–76.

[86] Wisner B, Blaikie P, Cannon T, Davis I. At risk: natural hazards,people’s vulnerability and disasters. Routledge; 2004.

[87] Wisner B, Gaillard JC, Kelman I. Challenging risk – has the left footstepped forward? In: Wisner B, Gaillard JC, Kelman I, editors.Handbook of hazards and disaster risk reduction. Abingdon,Oxfordshire, U.K.: Routledge; 2012. p. 789–93 (Chapter 65).

[88] Wisner B, Gaillard JC, Kelman I. Framing disaster: theories andstories seeking to understand hazards, vulnerability and risk. In:Wisner B, Gaillard JC, Kelman I, editors. Handbook of hazards anddisaster risk reduction. Abingdon, Oxfordshire, U.K.: Routledge;2012. p. 18–33 (Chapter 3).

[89] Woodruff A. Samoa technical report – economic analysis of floodrisk reduction measures for the lower Vaisigano catchment area. EUEDF – SOPAC Project Report 69g Reducing Vulnerability of PacificACP States. SOPAC (Pacific Islands Applied Geosciences Commis-sion), Suva, Fiji; 2008 February.

[90] World Bank. Weather and climate services in Europe and CentralAsia. A regional review.Working Paper 151. Washington, D.C., U.S.A:The World Bank; 2008.





























Northumbria Research Linknrl.northumbria.ac.uk/17954/1/does_mitigation_save.pdf · Received 28 April 2014 Received in revised form 12 August 2014 Accepted 13 August 2014 Available

Documents